To determine the incidence of hospital-acquired (nosocomial) infection in pediatric long-term care facilities.
An 87-bed pediatric long-term care facility.
All patients receiving long-term care at Bloorview Children's Hospital during the study period.
Infection developed in 40.1% of patients (n = 456). The nosocomial infection rate per 1000 patient days (mean, 7.84) varied substantially, from 1.66 in May 1988 to 16.37 day in April, 1989. The proportional frequencies of infections were as follows: respiratory, 41.6% (37.0% upper, 4.6% lower); urinary tract, 31.0%; skin, 15.6% (gastric tube site 5.0%, other 10.6%), eyes, 6.4%; gastrointestinal, 3.5%; and other, 1.5%. Of those infections for which an organism was recovered (48.5%), pathogens included Escherichia coli (22.5%), Enterococcus (14.8%), Staphylococcus (14.8%), Streptococcus (11.2%), Klebsiella (10.5%), Pseudomonas (10.1%), Proteus (4.3%), yeast (4.3%), Salmonella (0.7%), Clostridium difficile (0.4%), and other (6.2%).
The incidence and nature of infections in pediatric long-term care facilities differs from those in acute care facilities. Physicians should become familiar with the infection rates in the populations whom they treat. Control requires compliance with currently recognized effective strategies as well as innovative practical approaches to respiratory disease. Behavioral problems related to frequent clean, intermittent catheterization in young adults need to be addressed.
During a 4-year period 4684 nosocomial infections occurred in a university pediatric hospital which admitted 78,120 patients (nosocomial infection rate (NIR) = 6.0). NIR varied from 0.17 to 14.0 on different wards or services; the highest rates (greater than or equal to 5.6) were found in the Neonatal Intensive Care Unit, infant neurosurgery, hematology/oncology, neonatal surgery, cardiology/cardiovascular surgery, Pediatric Intensive Care Unit and infant/toddler medicine areas. Infections were most common in patients less than or equal to 23 months (NIR = 11.5), were less common in the 2- to 4-year age group (NIR = 3.6) and occurred least frequently in patients greater than or equal to 5 years (NIR = 2.6). The median day of onset of infections was 15.3 days. The proportional frequencies of infections were: 35% gastrointestinal; 21% bacteremia; 16% respiratory (10% upper, 6% lower); 7% postoperative wound; 6% urinary tract; 5% skin (32% of these skin infections were related to intravascular lines); 5% eye; 3% cerebrospinal fluid; and 2% other. A similar proportional frequency of 379 infections in patients hospitalized for more than 100 days was observed. The etiologic agents were Gram-positive bacteria (50%), viruses (23%), Gram-negative bacteria (18%), fungi (4%) and mixed/other (5%).
Of 145 patients admitted to our hospital because of encephalitis-like illness, 50 patients hospitalized for > or =72 hours underwent standardized microbiological investigations. A confirmed or probable etiologic agent was identified in 20 cases (40%), including Mycoplasma pneumoniae (9 cases). M. pneumoniae and enterovirus (2), herpes simplex virus (4), Epstein-Barr virus (1), human herpes-virus 6 (HHV-6) (1), HHV-6 and influenza virus type A (1), influenza virus type A (1), and Powassan virus (1). In 13 cases (26%), a possible pathogen was identified, including M. pneumoniae in nine cases. Presenting features included fever (80% of patients), seizures (78%), focal neurological findings (78%), and decreased consciousness (47%). The frequency of findings at the time of admission vs. later in hospitalization was as follows: pleocytosis, 59% vs. 63%; electroencephalogram abnormalities, 87% vs. 96%; and neuroimaging abnormalities, 37% vs. 69%, respectively. The outcomes at the time of discharge were as follows: normal results of physical examination, 32% (16) of the patients; death, 2% (1); motor difficulties, 26% (13); global neurological deficits, 16% (severe, 6; mild, 2); mental status changes, 14% (7); visual defects, 8% (4); and hearing impairment, 2% (1).
For determination of the incidence of viral-associated diarrhea after admission to a pediatric hospital, all patients admitted to general pediatrics, cardiology, and neurosurgery wards without diarrhea between January 1 and July 31, 1985 were followed 5 days per week for presence of diarrhea, etiologic agent, and possible risk factors. A total of 1,530 patients were followed for 3,642 days. Of these patients, 69 developed 80 nosocomial diarrhea episodes after 72 hours in hospital for a nosocomial diarrhea rate of 4.5 infected children per 100 admissions. Of 358 patients with an infected roommate, 37 (10.3%) developed nosocomial diarrhea. Etiologic agents recognized included rotavirus (43%), calicivirus (16%), astrovirus (14%), minreovirus (12%), adenovirus (8%), Salmonella sp. (4%), and parvo/picornavirus (3%). The nosocomial diarrhea rate by age was: 0-11 months, 8.8%; 12-35 months, 3.6%; and 36 months or more, 0.6%. The rate by length of stay was: 3-7 days, 8.4%; 8-14 days, 10.4%; 15-21 days, 7.9%; and 22 days or more, 8.8%, and by number of roommates/1,000 patient-days it was: 0-1, 15.7; 2 to 3, 27.7; and 4 or more, 45.2. Patients who acquired diarrhea were more likely to be diapered (9.6% vs. 1.8%, p less than 0.001). Playroom use was not significantly different in the two groups. A total of 64 patients developed diarrhea within 72 hours of admission (community diarrhea rate = 4.2). Nosocomial viral-associated diarrhea is almost exclusively a disease of diapered children less than age 36 months and occurs at any time during hospital stay. It is more common in multibed rooms, but does occur in single-bed rooms.
In a prospective 12-month study at a university-affiliated pediatric hospital, isolation usage was quantitated by ward/service, season, isolation category and type of infection (community-acquired vs nosocomial). Such information may be helpful in designing hospitals, recognizing time utilization of the pediatric infection control nurse, and defining educational and isolation needs. Hospitals with multiple bed rooms and inadequate numbers of single rooms may be unable to meet current federal isolation guidelines. The mean number of isolation days was 153 per 1000 patient days or 15.3% of bed days used. This ranged from 18.5% on the infant/toddler/preschool medical ward to 2.8% on child/teenage orthopedic surgery. Isolation requirements vary seasonally and rose to 32% in winter on one ward. Proportional frequencies of isolation category included enteric--29%, protective--28%, strict--16%, barrier (contact)--10%, multiply resistant organism (MRO)--8%, wound--5%, pregnant women (careful handwashing)--3%, blood and body fluid precautions--1%. Isolation of patients with and contacts of nosocomial infections account for 32% of isolation usage. During one third of the 365-day year, the hospital is unable to provide adequate numbers of single rooms for one to 20 patients.
Substantial progress has been made in measuring the burden of nosocomial infection in pediatric patients, particularly in certain populations (e.g., critical care, immunocompromised, chronic care, and patients with acquired immunodeficiency syndrome) and after certain procedures (e.g., central catheter lines and open-sternum cardiovascular surgery). Preventive measures, such as the use of goggles, gowns, and gloves, have been subjected to new and additional study. The following report is a summary of recent progress. A review of factors responsible for infection in various patient care populations and settings and recommendations for control are presented.
After two cases of acute hepatitis B infection occurred in phlebotomists at The Hospital for Sick Children in 1985, a seroprevalence survey of hepatitis B virus markers was undertaken. Directors in high-risk areas were advised by phone and memorandum to screen employees. Participation was entirely voluntary, and employees who did not respond were contacted six weeks after initial notification. Information obtained from each participant through a self-administered questionnaire included age, duration of employment in current pediatric occupation, history of blood transfusions, immune globulin prophylaxis, needlestick injury, and country of birth (North America/United Kingdom or other). Sera identified by code were tested by radioimmunoassay. Interactions were analyzed by using a multiple logistic regression model. A total of 10% of the personnel in high-risk areas, in which there was frequent exposure to blood or blood products, had hepatitis B markers in their blood, compared with 2% who did not have this exposure. Birthplace and occupation have independently significant effects on the likelihood of having hepatitis B markers. There appears to be an increased risk to employees in pediatric units, depending on the patient (and parent) population being served, although the risk may be lower than in adult hospitals.